Proposal Summary/Abstract Polyketides are among the most diverse and complex natural products. The antibacterial, antifungal, anticancer, antiparasitic, and immunosuppressive properties of polyketides have made them invaluable pharmaceuticals and drug precursors. The emergence of drug-resistant microbes has increased the demand for new macrolide antibiotics derived from polyketides. Biosynthetic methods have the potential to rapidly generate therapeutic compounds in high yields with high chemo- and enantioselectivities. Previous attempts to re-engineer the multi-enzyme complexes that generate polyketides, known as Type I polyketide synthases (PKSs), have focused on modifying, exchanging, deleting, or duplicating modules. Although some novel polyketide products have been obtained from these efforts, engineered PKSs are often inactive or have greatly attenuated catalytic activity. This is partially due to a poor understanding of how 1) individual domains within modules function and 2) what structural features alter the reactivity of domains. To address these questions, we will modify PKS thioesterase (TE) domains, which catalyze the hydrolysis or cyclization of polyketide intermediates. This work will be driven by the hypothesis that the fundamental insights gathered from structural, activity, and mutagenic studies of TE domains will enable the development of cyclases with (more) predictable selectivities and serve as a foundation for future PKS engineering efforts. Building upon the previous work of Sherman group, we will attempt to 1) reveal the structural features that enable TE domains to selectively generate macrolides of different sizes, 2) generate a new class of head- to-tail cyclases with applications to organic synthesis, and 3) synthesize new macrolides with potential therapeutic applications. To achieve these aims, we will identify novel TE domains that can catalyze two different types of selective macrolactonization reactions. The substrates required for these reactions will be derived from macrolide aglycones and fully synthetic fragments. We will address related objectives by identifying and expressing two categories of TE?s: PikTE homologues that can form either 12- or 14-membered rings and TEs that catalyze regioselective macrolactonization. A semi-rational mutagenesis effort and a campaign of directed evolution via random mutagenesis will alter the activity of the wild-type TEs and provide information about the structural features that alter the selectivity of macrolactonization. The new macrolides obtained from our studies will be glycosylated and evaluated for biological activity. In collaboration with the Smith group at UM, crystal structures of novel TEs will be obtained with covalently tethered substrates. Calculations performed by members of the Houk group at UCLA can be used to determine the relative energetic barriers for the hydrolysis of polyketides and cyclization of polyketides at various positions. Previously unknown structure-activity relationships will be elucidated, and the regions of TEs that alter the selectivity of macrolactonization will be identified.